The overall objective of this research is to trace the course of individual differences in adaptation throughout childhood. The present research will extend longitudinal assessments of 190 children and their families through the middle childhood years. Assessments of these subjects began during the third trimester of pregnancy and continued through infancy, the toddler and preschool years and into early elementary school. Available data are extensive, including prenatal and newborn variables, early interaction, ongoing family circumstances, cognitive development, socioemotional development, and temperament. The specific focus of the present research concerns social relationships and social competence in middle childhood. Peer relationships, sibling relationships, and child-teacher relationships, as well as parent-child relationships, will be assessed. In addition to examining relations among these domains of social functioning and various contemporary correlates (family life stress, parental relationship stability, etc.), antecedents of individual differences in social functioning will be explored. The role of early parent-child relationships, child temperament, cognitive variables, etc., can be examined singly or in combination as they forecast particular features of relationships with peers and others. Thus, in addition to advancing the science of social relationships with peers in middle childhood, the present research will provide important data on the nature and origins of individual differences in social competence with peers and others in the school age child's social world. It is known that the adequacy, or inadequacy, of such relationships strongly predicts adult adjustment and psychopathology. A comprehensive understanding of such individual differences, including their origins, may be a key to efforts at primary prevention. A broad spectrum of methods and analytic procedures, all previously developed in earlier phases of this research, will be used. In brief, multiple data sources (teachers, peers, parents, observers and subjects) and multiple data generating techniques (ratings, Q-sorts, interviews, coded observations) are used. Validity is established through a convergence of sources and methods.